Control and Optimization for Observations of Systems Governed by Controlled Partial Differential Equations

This project was focused on fundamental issues of modeling and control of flexible structures. Particular problems considered were: patterns of transmission zeroes, sensor placement, robust control, and high-performance control using iterative learning approach. During the grant period a flexible beam control devise for experimental purpose was also developed. With most of the work carried out to the extent of laboratory implementation, the following results were concluded: 1) For flexible structure control the location of transmission zeroes in the mathematical models critically depend on the location of sensors. For flexible beams the movement of zeroes were mapped out for control design; 2) For flexible structures a class of new compensators called generalized lead/lag compensators were developed. Their implementation is very simple with delay elements; and 3) Optimal sensor placement can be based on the robustness of the dynamic observer to parameter uncertainty. A design procedure was developed for this purpose.